Jump to content
UltravioletPhotography

[Filter Test] 340/10 #4 Edmund Hard-coated 340/10 Rear-mount Experiment


Andrea B.

Recommended Posts

See Also:

[Filter Test 340/10 #1] Gotta work out a problem or two, but here is a 340/10 result anyway

[Filter Test 340/10 #2] Cured the flare but it's ring around the posy here

[Filter Test 340/10 #3] Trying again with the 340/10

[Filter Test 340/10 #4] here

[Filter Test 340/10 #5] Edmund Hard-coated 340/10, Sunflower, Sun

[Filter Test] 340/10 #6 Progress was made, finally. UPDATE: Progress was then lost.


 

 

 

Gear:

  • Narrowband, hard-coated, UV-pass Filter: Edmund 340/10, product number 65190.
  • Camera: Nikon D610 conversion by MaxMax
  • Lens: UV-Nikkor 105/4.5

.

My initial attempts to use this mirrored filter were disasters due to the extreme flare and reflection artifacts from its two shiny silver surfaces. Today I finally put together some parts to try rear-mounting this filter in hopes of supressing the artifacts. And then it began to rain. I shot anyway just to see what would happen. Surprisingly perhaps, I did get a record of whatever was out there between 330-350 nm.

 

Rear mount of filter in order from camera mount to lens mount:

  • Nikon K2 F-mount to 52mm ring
  • Edmund filter holder base
  • 340/10 hard-coated filter inside base
  • Nikon K3 52mm to F-mount ring

.

Of course all this extension cost me infinity focus on the UV-Nikkor, but that was OK because I simply wanted to see the raw record of the capture and also verify that I didn't see any dark central spots or rings. (Is a dark central spot called a cold spot?)

 

The photo I'm going to show you was intentionally blurred. The photo is of some leaves, a cement porch step and a brick sidewalk.

 

Raw Composite from Raw Digger. Color samples from PS Elements 11.

I sampled 4 areas in the raw photo. Each sample's Hue/Saturation/Brightness and RGB triplet are recorded. The saturation and brightness were boosted to 100% for the color circles associated with each sample. I do that just to help me to better "see" what the colors are so I can possibly name them. YMMV on that.

Small typo: ISP should be ISO, of course.

As a side note I observe that using f/4.5 and ISO 3200 for 10 long seconds in a UV photo made with this high transmission filter and a UV-Nikkor should be telling you that there really wasn't very much light available. I am looking forward to a repeat of this experiment on a nice sunny day.

d610_uvEdmund340x10_rainyOutdoor_20180610wf_10381rawComp.jpg

 

 

Red Channel from Raw Digger

d610_uvEdmund340x10_rainyOutdoor_20180610wf_10381redChan.jpg

 

Green Channel from Raw Digger

d610_uvEdmund340x10_rainyOutdoor_20180610wf_10381greenChan.jpg

 

Blue Channel from Raw Digger

d610_uvEdmund340x10_rainyOutdoor_20180610wf_10381blueChan.jpg

 

 

Auto WB in Raw Colours from Raw Digger

No saturation or contrast applied.

d610_uvEdmund340x10_rainyOutdoor_20180610wf_10381rawRender.jpg

 

Click-away the Magenta WB from Photo Ninja

Default contrast and Plain 65 Saturation. Here we see the familiar blues & yellows.

d610_uvEdmund340x10_rainyOutdoor_20180610wf_10381pn.jpg

Link to comment

Weeeellllll.....I did not intentionally set out to blur things. :lol:

 

It was just a combination of factors. The extension on the UV-Nikkor ruined infinity focus bigtime. I cannot see through this filter at all in order to attempt an actual closer focus. And I have no auxiliary UV-Led to help with this. It was drizzling so I couldn't really be in slow mode for a patient set up outdoors or use a broader flash. Thus I just tried to shoot at the screened porch door boundary into the best light available for capturing some raw colors and just let focus go. In other words, mostly I was just fooling around.

 

I'll get back to the experiment when the weather clears up. I really do want to see whether I can make an actual narrowband photograph with this difficult filter by rear-mounting it in hopes of squelching the wild reflections and dark spot. It initially looks like this might have worked. But maybe things could go askew again once I've managed to acquire some focus? Don't know, we shall see!

 

Here are links to the two previous 340/10 tests with the awful ring and flare artifacts.

 

[Filter Test 340/10 #1] Gotta work out a problem or two, but here is a 340/10 result anyway

 

[Filter Test 340/10 #2] Cured the flare but it's ring around the posy here

 

[Filter Test 340/10 #3] Trying again with the 340/10

 

In that last experiment I also looked at the raw colors. So I duly note that they're in the same neightborhood as I got here.

Link to comment

It took about 30 minutes to find the focus for this next photograph. It was like going back to the dark ages of no Live View. The UV-Nikkor infinity focus moved to about 30 centimeters under the extensions needed for rear-mounting the Edmund 340/10. What a pain !!

 

The good news is that I was able to make a 340/10 photo without flare and without reflection rings for the first time. I do hope the success can be repeated in the big sunshiny outdoors. If not, then I'm going to repurpose this pesky filter as a purse mirror for touching up lipstick - the most expensive purse mirror ever it will be!

 

The bad news is that this photograph did not want to be white balanced.

 

You might kindly ignore the fact that this successful photograph has no meaningful content other than its color? I'm just trying to refine methodology before I try to shoot something beautiful which will fit into the frame when only 30 centimeters away from the sensor plane.

 

 

Raw composite

340/10 on UV-Nikkor on D610mod

f/5.6 for 20" @ ISO-100 with 3 flashes of the SB-14mod.

The Exif in the file is incorrect for aperture because the extension tubing squelches the lens chip communication.

d610_uvEdmund340x10_sb14indoor_20180610wf_10451rawComp.jpg

 

 

Raw Digger auto WB in raw colors

That is 99% reflective Spectralon in the background under the piece of paper. It did not turn white. But then I don't know how Raw Digger's auto WB works. I suspect it is a kind of averaging given the next result I got.

d610_uvEdmund340x10_sb14indoor_20180610wf_10451rawRender.jpg

 

Capture NX2 marquee WB on the Spectralon

The Spectralon would not turn white even in Nikon's converter using their WB marquee tool. I enlarged the marquee and made an average white balance which turned out to look about the same as the marquee attempt on the Spectralon only. It is rather unusual (at least for me) to see a WB failure in NX2.

d610_uvEdmund340x10_sb14indoor_20180610wf_10451nx2AveWB.jpg

 

Photo Ninja WB drag on the Spectralon

Finally, I got almost a white Spectralon thanks to the Ninja. But it took two tries (!) and the blue became very intense. The bandpass is so narrow on this Edmund 340/10 filter that I am not surprised about the white balance difficulties. For white balance tools to work best, I think you need some wider range on the color wheel than this indoor shot provides. Else you are just attempting to drag apart a narrow wedge of the color wheel. (Clark has made a similar observation in the past about narrowband filters and white balance.) But anyway, I first white balanced the raw file in PN, then I re-white-balanced that result again in PN. The first time through PN left the file with the very slightest yellow tint.

d610_uvEdmund340x10_sb14indoor_20180610wf_10451pnpn.jpg

 

Raw Color Samples

d610_uvEdmund340x10_sb14indoor_20180610wf_10451rawComp01.jpg

Link to comment

10nm is not much to white balance from.

340BP10 12.5mm filter mounted on the front of Kuribayashi 35mm, @f/5.6. Outdoor sunshine.

Indeed, very little difference between in camera WB, and out of camera NX2 Marquee WB.

post-87-0-77275600-1528682910.jpg

 

NX2 WB

post-87-0-05039800-1528682954.jpg

Link to comment
So this is an average white balance given that NX2 won't WB a 340/10 properly?
Link to comment

I don't understand what you are asking.

The lower shot above is NX2 marquee WB, not exactly full screen because this is a crop to cut off some of the vignetting, the black has some blue noise like your examples.

It is hard to see the blue noise in the example above, so here is a close up from the lower left corner.

 

Out of camera

post-87-0-88999400-1528685842.jpg

 

NX2 Marquee WB

post-87-0-36645500-1528685859.jpg

Link to comment
Andy Perrin

The problem is similar (but maybe even more extreme) than when using an LED torch. Some of those blue and yellows look kind of nice, though? Perhaps it will have some arty use.

 

Or you could use it as one of the channels for one of those "three band" UV photos that were being discussed in the other thread?

Link to comment

It is not a White Balance if the WB marquee or dropper is not placed on some area in the photo which should be white (or light grey). If the WB marquee or dropper is used over the entire photo, or over most of the photo, then you have made an Average White Balance. An Average White Balance may not turn the supposed-to-be-white subject white. The difference between WB and Average WB can be quite striking as shown in my Spectralon & Paper photo above. I'm not entirely sure why Average White Balance is considered white balance at all. Maybe because the AWB is made with a WB tool even though the result doesn't necessarily produce a white or neutral.

 

The additional complication is that sometimes the marquee tool will not work even though it is placed over a supposed-to-be-white subject. That is because NX2 cannot reach a low enough temperature in its WB tool for certain unusual situations -- like this narrowband 340/10. The marquee tool would not make that Spectralon white like it should be. So I had to switch to Photo Ninja. And even in Photo Ninja, I had to make two passes to get an actual white on that Spectralon.

Link to comment

OK, Marquee = average then.

All I was saying is that 10nm is not much to work with for a white balance, and you probably will not get much of a range from that bandwidth.

Would you like a 10nm example from somewhere in the the visual range?

Link to comment

yes!! That would be cool. Thank you.

 

And totally agree about 10mm not being much to work with as you and Clark have observed. It seems to force the white balance. However, I do think the average WB produced is rather pretty with its lime yellow and blue. :)

 

No sunshine today, so I cannot continue my 340/10 rear-mount experiment. :(

Link to comment

Cloudy/rainy here today, and I want to do the next test with dandelion or the like instead of daisy, better UV test target.

When I get to it, I will do a visual range 10nm BP shot also.

Your NX2 looks quite similar to my NX2 with the 340BP10.

The thing is, that pushes the WB so much that it starts to make things noisy with the blue.

Hard to extract white from the Bayer filters seeing just a narrow band, perhaps.

However, your Ninja shot get's white. But depending on what and how we use the shot, do we really want white, if we (say) desaturate for a stack... I don't know,

just thinking.

Question: Lets take the scenario where I use 3 different UV BP filters to construct a RBG composite.

What do you think is the best processing of each BP shot?

Convert each to monochrome, perhaps? White balance each first? Anything else?

What are your thoughts?

Link to comment
I'm not able to answer that question currently because I do not have a channel app (as in Photoshop) to test with. I have to kludge my RGB composites and they may not be correctly done. Post this as a separate question in its own topic and perhaps some good answers will be posted. :D
Link to comment
Well it is important and I wish I had a good answer !!
Link to comment

For now, here is a shot using a 330AF20, twice the bandwidth of a 10nm wide bandpass filter.

This shot is WB on the virgin PTFE sheet material that the Sparticle housing is made from, and yet the dark areas still have blue noise.

So I am pretty sure there is not much Bayer color range from such narrow bands. I will try a flower soon.

post-87-0-49111100-1528779662.jpg

Link to comment

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
×
×
  • Create New...